Chapter Contents |
Previous |
Next |

EWMACHART Statement |

See MACEW2 in the SAS/QC Sample Library |

By default,
the EWMACHART statement estimates the process mean ()and standard deviation ()from the data. This is illustrated
in the "Getting Started"
section of this chapter. However,
there are applications in which standard values
( and ) are available
based, for instance, on previous experience or
extensive sampling. You can specify these values with the MU0=
and SIGMA0= options.
For example, suppose it is known that the metal clip manufacturing
process (introduced
in "Creating EWMA Charts from Raw Data" )
has a mean of
15 and standard deviation of 0.2. The following
statements specify these standard values:

title 'Specifying Standard Process Mean and Standard Deviation'; symbol v=dot c=salmon; proc macontrol data=clips1; ewmachart gap*day / mu0 = 15 sigma0 = 0.2 weight = 0.3 xsymbol = mu0 cframe = vibg cinfill = ligr coutfill = yellow cconnect = salmon; run;

The XSYMBOL= option specifies the label for the central line. The resulting chart is shown in Output 20.1.1.

The central line and control limits are determined using and (see the equations in Table 20.19). Output 20.1.1 indicates that the process is out-of-control, since the moving averages for DAY=17, DAY=19, and DAY=20 lie below the lower control limit.

You can also specify and with the
variables _MEAN_ and _STDDEV_ in a LIMITS= data set,
as illustrated by the following statements:

data cliplim; length _var_ _subgrp_ _type_ $8; _var_ = 'gap'; _subgrp_ = 'day'; _type_ = 'STANDARD'; _limitn_ = 5; _mean_ = 15; _stddev_ = 0.2; _weight_ = 0.3; proc macontrol data=clips1 limits=cliplim; ewmachart gap*day / xsymbol=mu0; run;

The variable _WEIGHT_ is required, and its value provides the
weight parameter used to compute the EWMAs.
The variables _VAR_ and _SUBGRP_ are also required, and their
values must
match the *process* and *subgroup-variable*, respectively,
specified in the EWMACHART statement.
The bookkeeping
variable _TYPE_ is not required, but it is recommended to
indicate that the variables _MEAN_ and _STDDEV_ provide
standard values rather than estimated values.

The resulting chart (not shown here) is identical to the one shown in Output 20.1.1.

Chapter Contents |
Previous |
Next |
Top |

Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.